| 1 | import codecs, encodings |
|---|
| 2 | #orginal code by Paul Prescod, submitted to ASPN |
|---|
| 3 | |
|---|
| 4 | """Caller will hand this library a buffer and ask it to either convert |
|---|
| 5 | it or auto-detect the type.""" |
|---|
| 6 | |
|---|
| 7 | # None represents a potentially variable byte. "##" in the XML spec... |
|---|
| 8 | autodetect_dict={ # bytepattern : ("name", |
|---|
| 9 | (0x00, 0x00, 0xFE, 0xFF) : ("ucs4_be"), |
|---|
| 10 | (0xFF, 0xFE, 0x00, 0x00) : ("ucs4_le"), |
|---|
| 11 | (0xFE, 0xFF, None, None) : ("utf_16_be"), |
|---|
| 12 | (0xFF, 0xFE, None, None) : ("utf_16_le"), |
|---|
| 13 | (0x00, 0x3C, 0x00, 0x3F) : ("utf_16_be"), |
|---|
| 14 | (0x3C, 0x00, 0x3F, 0x00) : ("utf_16_le"), |
|---|
| 15 | (0x3C, 0x3F, 0x78, 0x6D): ("utf_8"), |
|---|
| 16 | (0x4C, 0x6F, 0xA7, 0x94): ("EBCDIC"), |
|---|
| 17 | } |
|---|
| 18 | |
|---|
| 19 | def autoDetectXMLEncoding(buffer): |
|---|
| 20 | """ buffer -> encoding_name |
|---|
| 21 | The buffer should be at least 4 bytes long. |
|---|
| 22 | Returns None if encoding cannot be detected. |
|---|
| 23 | Note that encoding_name might not have an installed |
|---|
| 24 | decoder (e.g. EBCDIC) |
|---|
| 25 | """ |
|---|
| 26 | # a more efficient implementation would not decode the whole |
|---|
| 27 | # buffer at once but otherwise we'd have to decode a character at |
|---|
| 28 | # a time looking for the quote character...that's a pain |
|---|
| 29 | |
|---|
| 30 | #truncate long buffer: |
|---|
| 31 | buffer = buffer[0:4] |
|---|
| 32 | encoding = "utf_8" # according to the XML spec, this is the default |
|---|
| 33 | # this code successively tries to refine the default |
|---|
| 34 | # whenever it fails to refine, it falls back to |
|---|
| 35 | # the last place encoding was set. |
|---|
| 36 | bytes = (byte1, byte2, byte3, byte4) = tuple(map(ord, buffer)) |
|---|
| 37 | |
|---|
| 38 | enc_info = autodetect_dict.get(bytes, None) |
|---|
| 39 | |
|---|
| 40 | if not enc_info: # try autodetection again removing potentially |
|---|
| 41 | # variable bytes |
|---|
| 42 | bytes = (byte1, byte2, None, None) |
|---|
| 43 | enc_info = autodetect_dict.get(bytes) |
|---|
| 44 | |
|---|
| 45 | if enc_info: |
|---|
| 46 | encoding = enc_info # we've got a guess... these are |
|---|
| 47 | #the new defaults |
|---|
| 48 | |
|---|
| 49 | # try to find a more precise encoding using xml declaration |
|---|
| 50 | secret_decoder_ring = codecs.lookup(encoding)[1] |
|---|
| 51 | (decoded,length) = secret_decoder_ring(buffer) |
|---|
| 52 | first_line = decoded.split("\n")[0] |
|---|
| 53 | if first_line and first_line.startswith(u"<?xml"): |
|---|
| 54 | encoding_pos = first_line.find(u"encoding") |
|---|
| 55 | if encoding_pos!=-1: |
|---|
| 56 | # look for double quote |
|---|
| 57 | quote_pos=first_line.find('"', encoding_pos) |
|---|
| 58 | |
|---|
| 59 | if quote_pos==-1: # look for single quote |
|---|
| 60 | quote_pos=first_line.find("'", encoding_pos) |
|---|
| 61 | |
|---|
| 62 | if quote_pos>-1: |
|---|
| 63 | quote_char,rest=(first_line[quote_pos], |
|---|
| 64 | first_line[quote_pos+1:]) |
|---|
| 65 | encoding=rest[:rest.find(quote_char)] |
|---|
| 66 | return encoding |
|---|
